Inês Brito Pereira's master's thesis, under the supervision of Luís Bernardo, our researcher and Professor at FCT UNOVA Lisbon, won the “APDC Telecommunications Best Thesis Award” in 2024.

The thesis, entitled “Algorithms for Physical Layer Authentication”, was carried out within the scope of the integrated master’s degree in electrical engineering and computers at the Faculty of Science and Technology of Universidade Nova de Lisboa. 

Authentication in wireless networks is crucial due to their vulnerability to security threats. To enhance security new methods are being explored, such as Radio Frequency Fingerprinting. This approach relies on the fact that it is impossible to have two devices with the same physical characteristics, allowing it to extract unique features from a device's signal for authentication. In this study, Inês focused on how to use self-generated features extracted from a signal by contractive autoencoders to implement device authentication at the physical layer under the communication protocol IEEE 802.15.4. For this purpose, the performance of different types of neural networks with different latent state dimensions was studied, as well as two classifiers: a Multi-layer Perceptron and a Random Forest. In addition, the GNU Radio software was used to process the samples and extract the signals’ Synchronization Header from the wireless sensors required to create the database to develop and test the algorithms.

APDC announced at its 33rd Digital Business Congress the winners of the 4th edition of the Best Thesis Award, in the media, IT, and telecommunications categories. It is developed in partnership with the Consortium of Engineering Schools and Academies participating in the UPskill Program, and Axians Portugal sponsors.

Thanks to APDC for the video.

Watch the conference video (the interval 3:39:40-3:41:05):